MODELING METHODOLOGY BASED ON ARTIFICIAL IMMUNE SYSTEM ALGORITHM AND NEURAL NETWORKS APPLIED TO REMOVAL OF HEAVY METALS FROM RESIDUAL WATERS

摘要

Artificial Immune Systems (AIS) and Neural Networks (NN) are biologically inspired methods that, due to their flexibility and performance have a huge potential for modeling complex processes from chemical engineering field. In this paper, an algorithm based on Clonal Selection (CS) principle of AIS acts as an optimizer for a neural network, the proposed methodology being named CS-NN. The optimal neural model is applied to simulate the removal of heavy metals from residual waters. In order to determine the performance of CS-NN, a series of simulations based on exprimental data were performed. The results obtained indicated that the methdology is able to determine a good model, even when using data with high measurement errors.